Nov 3, 2026 · 1h 6m
Dan Niles warns AI’s financing boom could unwind
Big Tech Selloff: Dan Niles on What Happens Next
The episode tests whether today’s AI selloff is a temporary correction or an early warning of a broader bust driven by scarce capital, infrastructure limits and macroeconomic risk.
- 1AI adoption remains a long-term growth story, but financing pressures could trigger a severe near-term drawdown.
- 2Anthropic’s IPO could become a crucial test of investor confidence, while OpenAI faces pressure between enterprise rivals and consumer platforms.
- 3Agentic AI may reshape software, advertising and chip demand, making selectivity more important as infrastructure and policy risks mount.
Don't miss
Niles explains why a failed or disappointing Anthropic IPO could signal that the AI trade has already peaked.
The brief
Dan Niles revisits his forecast for a 30% to 50% AI correction, arguing that repriced financing and circular funding echo the late-1990s internet bubble.
Anthropic’s planned IPO becomes the episode’s market test: its enterprise focus may offer resilience, while OpenAI looks more exposed between enterprise rivals and consumer platforms.
Niles remains bullish on enterprise adoption and agentic workflows, which could change staffing, productivity, software spending and the balance between CPUs and GPUs.
The discussion weighs Google, Meta, Microsoft, Nvidia and Apple against different AI pressures, from search economics and chip competition to cash flow and infrastructure.
Niles’s central discipline is tactical rather than ideological: keep the long-term AI thesis, but reduce exposure when capital tightens, politics intervenes or macro risks rise.
What was said on this episode
34 statements · 16 positive · 15 negative · 3 neutral
Anthropic is better positioned than OpenAI.
“Anthropic is the better company.”
Listen at 0:26
OpenAI will be squeezed between enterprise and consumer competitors.
“I think OpenAI has got a huge problem because they're stuck between anthropic and enterprise and Google and consumer. And so I think they're going to get squeezed.”
Listen at 0:38
AI investment will eventually experience a major bust.
“which then means that you're eventually going to have a big bust.”
Listen at 5:06
The current AI cycle resembles 1997–1998 more than the internet bubble peak.
“And I think it's more like 97, 98 with the internet where you had...”
Listen at 5:41
Agentic AI requires 10–100 times more tokens than chat-based AI.
“Agentic AI, and I'm sure you've talked about Muse on some of your podcasts from Meta that just came out. or OpenClaw, which came out in January, Agenic AI requires 10 to 100 times more tokens.”
Listen at 6:09
Higher financing costs are making funding difficult for marginal AI companies.
“So you're seeing the cost of money go up tremendously, and the more marginal players are having a real tough time getting it.”
Listen at 8:42
Anthropic is best positioned to raise capital.
“And so they are the in the best position to raise money.”
Listen at 12:45
Anthropic’s IPO will likely occur at a lower valuation and raise less capital.
“I think the deal will happen, I think it may happen at a much lower valuation, they might be able to raise less money than what people thought.”
Listen at 12:57
Anthropic’s IPO will be completed.
“I think it will get done.”
Listen at 14:17
Microsoft has 450 million Microsoft 365 licenses.
“Microsoft has 450 million Microsoft 365 licenses.”
Listen at 15:54
Enterprises will pay for AI services.
“But enterprises are going to pay for this stuff.”
Listen at 21:42
Meta makes agentic AI easier for users through Muse.
“And so I think Meta, with the fact that they got out what I call an agent for the rest of us, like OpenClaw, which was formalized on January 30th, right? You had to buy a Mac Mini. You have to set it all up. Meta makes it easy with Muse.”
Listen at 23:42
Meta is well positioned to monetize agentic AI.
“And Meta is in a pretty good spot to potentially monetize that.”
Listen at 24:02
Anthropic may survive because of its enterprise strength.
“my belief is longer term i think anthropic might make it out of this okay because they are such a good enterprise tool”
Listen at 31:52
OpenAI will be squeezed between enterprise and consumer competitors.
“i think open ai has got a huge problem because they're stuck between anthropic and enterprise and Google and consumer. And so I think they're going to get squeezed.”
Listen at 32:07
Agentic AI requires more CPUs relative to chat-based AI.
“Agents require more CPUs relative to chat-based AI.”
Listen at 37:23
Intel is well positioned to supply U.S. advanced semiconductor capacity.
“And I think Intel is very well positioned to do that.”
Listen at 38:10
Intel will announce foundry agreements with major hyperscalers and chipmakers next year.
“And I think you'll see them announce this next year agreements with some of the biggest hyperscalers on the planet, as well as semiconductor manufacturers, giving you a viable alternative to TSMC or Samsung. on the foundry side”
Listen at 38:14
Intel’s fundamental position should improve over the next two years.
“I do think next year, year after, it should improve their fundamental positioning, hopefully.”
Listen at 39:10
Apple’s December-quarter estimates will be revised downward.
“I think the December quarter numbers are too high right now and they'll have to come down a little bit.”
Listen at 41:18
Next year will be a strong year for Apple.
“I think next year will be a huge year for Apple”
Listen at 41:34
NVIDIA and ASICs will both perform well in AI infrastructure.
“And so I think NVIDIA will continue to do well. I think ASICs will do well also.”
Listen at 44:47
Long-term memory supply agreements will not protect companies during a downturn.
“And I think you're going to see the same thing in the memory space, where you've got some companies signing stuff out until 2030, 31. And I go, this is just... This is ridiculous. This is not going to save you when this rolls over”
Listen at 46:13
A broad group of S&P stocks may perform adequately due to seasonality.
“I think a broader group of S&P names may actually do okay just because of seasonality.”
Listen at 50:20
Investors should reduce exposure when they anticipate major drawdowns.
“you want to, because you don't want to get caught in a massive drawdown.”
Listen at 51:16
Investors should adapt positions as market facts change.
“And as the facts change, I change.”
Listen at 53:36
Niles is more negative on AI than a year ago because capital is scarcer.
“Am I more negative than I was a year ago? 100 percent because capital is more scarce.”
Listen at 54:36
Neocloud companies are likely to fail or deteriorate substantially.
“I think the neoclouds, they're toast.”
Listen at 55:12
Easy monetary policy causes capital misallocation.
“I don't like easy money. Easy money leads to misallocation of capital.”
Listen at 57:57
The Federal Reserve will raise rates in December and early next year.
“I would be beyond stunned if they didn't raise in December and keep raising to start the next year once the midterms are behind us.”
Listen at 58:20
Sustained oil-price increases preceded 10 of the last 12 recessions.
“10 of the last 12 recessions have been preceded by a sustained increase in oil prices.”
Listen at 1:00:55
The AI trade will eventually experience a severe bust.
“Do not delude yourself into thinking that we're not going to have a horrific bust on the other side of this.”
Listen at 1:02:50
AI adoption still has substantial growth ahead.
“I think we have a long way to go.”
Listen at 1:03:11
Investors should avoid outsized risk in the AI trade.
“don't take outside outsized risk”
Listen at 1:03:27
Statements are attributed to the speaker as said on the episode and reflect their view at the time, not PodLume's. They are not advice.
Weekly email
State of the Pod, every Monday
The week's most-discussed names and stories from podcasts, with the moments behind them. Free, five minutes, unsubscribe in one click.